MINING AND METALLURGY INSTITUTE BOR ISSN: 2334-8836 (Štampano izdanje)

UDK: 622 ISSN: 2406-1395 (Online)

UDK: 622.33:681.51(045)=111 doi: 10.5937/mmeb2002001V

Nenad Vušović*, Milica Vlahović**, Milenko Ljubojev***,

Miljan Vlahović****, Daniel Kržanović***

SOFTWARE SOLUTION FOR THE MINE SUBSIDENCE

PREDICTION BASED ON THE STOCHASTIC METHOD
INTEGRATED WITH THE GIS*****
Abstract

Mining induced the ground subsidence and consequently, damages on objects above excavation
works is a serious problem all over the world. In this paper, a new approach for subsidence and de-
formation prediction based on the stochastic Pataric-Stojanovic method is presented. Since the calcu-
lation using the equations of this method is very complex, the original computer program package
with the title MITSOUKO has been purposely designed. The MITSOUKO program enables calculat-
ing at any point of the land surface and representing the results owing to the possibility of their inte-
gration and further processing in the Geographical Information System (GIS). The components of the
GIS are used to realize the spatial analysis.
Keywords: coal mining; surface subsidence; stochastic prediction method; GIS, spatial analysis

1 INTRODUCTION

Mine surveying in coal mines all over
the world was initiated with the underground
coal mining- induced land subsidence and
building damage in the Ruhr Basin in the
middle of the 19th century. Based on the
obtained results, different methods for pre-
dicting subsidence and deformation parame-
ters, such as the empirical methods, the pro-
file function methods, the influence function
methods, and the void diffusion method
were developed [4].
Also, numerous theoretical studies were
performed using the stochastic method, the
elastic methods, and the visco-elastic method
as well as the finite-element methods, the
boundary element method, and the distinct
element method for the ground subsidence
and deformations prediction. Most of these
methods cover the horizontal and slightly
sloped coal seams. The methods for seams
with greater slopes are less reliable but with
a high professional improvisation. Further

*
University of Belgrade, Technical Faculty in Bor, Vojske Jugoslavije 12, Bor, Serbia,
Corresponding author's e-mail: nvusovic@tfbor.bg.ac.rs
**
University of Belgrade, Institute of Chemistry, Technology and Metallurgy, Karnegijeva 4, Belgrade,
Serbia
***
Mining and Metallurgy Institute Bor, Zeleni bulevar 35, Bor, Serbia
****
Public Company Nuclear Facilities of Serbia, Mike Petrovića Alasa 12-14, 11351 Vinča, Belgrade,
Serbia
*****
This work was financially supported by the Ministry of Education, Science and Technological
Development of the Republic of Serbia (Grant Nos. 451-03-68/2020-14/200131, 451-03-
68/2020-14/200026 and MMI).

No. 1-2, 2020 1 Mining & Metallurgy Engineering Bor

more, the numerical methods were used for
subsidence modeling and calculating the
rock strata movement [3,4]. The shape and
position of subsidence trough in relation to
the excavated area depend on the geological
and technological conditions of underground
exploitation; for the horizontal coal seams, it
is symmetrical, while for inclined it becomes
asymmetrical. The available methods relia-
bly prognose both vertical and horizontal
subsidence components and offer deriving
other displacement factors of movement,
such as the tilt, curvature, and horizontal
strain.
Prediction of the final subsidence is sim-
pler compared with the displacements in-
duced by sequential mining processes. Also,
many factors affect and conduct moving of
the rock mass thus making the behavior of
overlying strata complicated. Therefore,
different idealized media were used for the
subsidence prediction in the middle of the
20th century. The stochastic theory of
ground movements proved to be a universal
method for subsidence prediction [1].
The first application of the stochastic
model was in the soil mechanics for sand,
gravel and severely cracked ground. Pokrov-
sky (1929) concluded that the change in
pressure due to the concentrated force at the
horizontal plane as a function of depth can
be represented by the standard Gaussian
distribution curve, also known as a normal
distribution curve. According to J. Litwin-
iszyn, with a presumption that the rock mass
is a stochastic medium, the stochastic meth-
od interprets the displacement of rock massif
related to mine subsidence [2].
Geoinformation technologies offer the
management of spatial data with their inte-
gration in the Geographic information sys-
tem (GIS) [6,12]. The computer-based ana-
lytical methods that realistically simulate
subsidence processes are needed for the reli-
able design of mining layout to minimize the
underground excavation influence on the
land surface. Integrative modeling, interac
No. 1-2, 2020
tive spatial analysis and understanding dif-
ferent processes are supported by the GIS.
Based on the simulation of spatially distri-
buted and time-dependent subsidence pro-
cesses, innovative thematic maps containing
the land-surface properties can be designed
[7-11].
2 STOCHASTIC METHOD
BY PATARIC AND STOJANOVIC
The stochastic Pataric-Stojanovic
method applies the mathematical statistics
and assumes the multi-layered massif,
split by a series of cracks into a large
number of elements with movements hav-
ing stochastic character.
2.1 Movements in a massif with
horizontal seams
The environment is presented by the
symmetrically arranged elements with
similar dimensions. Such an area is not
real, but the assumption is statistically cor-
rect because the pressure change curve in a
homogeneous medium is symmetrical, thus
elements in the profile must be symmetrical
[1]. Owing to this symmetry, the force got
by one element is transmitted and equally
divided into two parts on which it relies
(Figure 1). For simplicity, a unit force is
observed, so if the surface layer is marked
as zero, the element in this layer receives
the entire force, the two elements in the
first layer, on which the element from the
zero layer relies, receive half of this force.
In the second layer, the edge elements re-
ceive a quarter, and the middle two quarters
of the force, etc. Following the process
further, it is noticed that the distribution of
forces obeys the binomial law. When
i, j  0 , the binomial distribution asymp-
totically tends to normal, then the pressure
changes according to the law of normal
distribution, and can be represented by a
Gaussian curve [2].
2 Mining & Metallurgy Engineering Bor
 Figure 1 Pressure distribution according to the stochastic model [2]

The mathematical model from Figure
1 can be used to study the plane subsid-
ence of a layered massif [3]. The subsid-
ence will be plane if a long area of con-
stant width ( 2a ) and height ( d ) has been
excavated. The assumption is that the rock
massif consists of the horizontal layers
divided by a series of cracks into equal,
symmetrically arranged elements (Figure
2).

Figure 2 Simmetrically arranged elements in horizontal seams divided by cracks [3]

When the element from the lowest layer
is removed, the elements in the higher layers
begin to move. Their movement is assumed
to be stochastic, so instead of studying the
displacement, the probability of the element
passing into the lower layer is determined. A
place that remains vacant in the lowest layer
can be occupied by one of the elements that
lie directly above it. Owing to symmetry, the
probabilities of these events are the same, so
their value equals ½. With further following
the probabilities for moving elements from
higher layers, the same regularity is valid
[2]. Since the subsidence is plane, two coor-
dinates are sufficient: an abscissa x (parallel
to the layers) and an elevation z (vertically
with the upward direction) to determine the
position of the elements in the massif.

No. 1-2, 2020 3 Mining & Metallurgy Engineering Bor

 For the boundary transition from a dis-
crete division to a continuous massif, the
starting point is the position of the elements
that are defined by the coordinates, ie:
F x, z  h  
(1)
 F x  b0 , z   F x  b0 , z 
1
2
where F x, z  is the function of subsidence
probability [3].
The assumption is that the excavated
area  2 has a rectangle shape with width
2a and thickness d , so if a coordinate
origin is placed in the intersection of rec-
tangle diagonals, it will be:
U x    F x, x, H , z  dx dz
2
Since the thickness d is negligible
with respect to the depth H of the seam,
instead of the integral, its mean value can
be taken without a significant error:
d /2
 F x, x, H , z  dz  d  F x, x, H ,0
d /2
Using the function from the Eq. (1) the
subsidence is:
t 1
 s2
 t   
2 seam dip. The direction of x -axis is chosen
e 2 ds
2 0
the subsidence is definitely calculated by
the formula [3]:
U x   U 0 X 0 x  (4)
where:
1  ax  a  x 
X 0 x      
2  n   n 
2.2 Movements in a massif with
inclined seams
In the case of inclined seams, the starting
assumption that the massif is divided by a
series of horizontal and vertical cracks into
elements of the same dimensions. Equation
(4) can only be applied to profiles along the
strike. However, the subsidence curve is not
(2) symmetrical by the seam dip, so there is
anisotropy unlike the process in a massif
with horizontal seams. The same mathemat-
ical model can be used assuming that the
elements, with still identical dimensions due
to homogeneity, are no longer symmetrically
arranged, but the center of each element is
shifted by the dip for a certain length e ,
whereby 0  e  b1 (Figure 3). The coordi-
nate origin is placed at an arbitrary point O
of the seam that is being excavated, the z -
axis is directed vertically upwards, the x -
axis is directed horizontally along the seam
strike, and the y -axis is in a direction of the
(3) in a way that the coordinate triedar is right.

Figure 3 Moving process in the massif with inclined seams [3]

No. 1-2, 2020 4 Mining & Metallurgy Engineering Bor

 The equation of the plane of the seam at strike and O by the seam dip.The sub-
sidence in a point M x, y, H  of the ter-
the distance t from the coordinate origin is:

t  z cos   y sin  (5) rain surface z  H of the undermined

massif based on the Eq. (7) will be a dou-
In this case, the place of the lowest re-
ble integral:
moved element D0 will be occupied by the
element D1 or by the element D2 , but U x, y    F x, x, y, y, z, z  dx dy dz
owing to the disturbed symmetry, the proba- 2
bilities of these events are not equal. Their (8)
values can be calculated as the resistances of
simply supported beam of span 2b1 loaded If the sides of the excavation panel  2
with a unit eccentric force, so the probability are 2a by the seam strike and 2l by the
is determined as follows [3]: seam dip, the coordinate origin is at the in-
tersection point of the rectangle diagonals
1 e  1 e  will be: a    a ;  b    b and using
   F x, y  b1 , t     
 2 2 b1   2 2b 1  the function:
F x, y  b1 , t   F x, y  e, t  h  x 1
 t2
 x   
2
(6) e 2 dt (9)
2 0
The massif is assumed to be anisotropic, a definite formula for calculating the sub-
therefore the subsidence curves in the pro- sidence during horizontal seam excava-
files by the seam strike are symmetrical, and tion is obtained [3].
asymetrical in the profiles by the seam stike
[3]. U x, y   U 0 X x Y  y  (10)
When deriving the equations, it is as-
sumed that the centers of elements remain in where:
the same vertical plane during subsidence. In
1   ax  a  x 
the general case, the center of element dur- X x     p   p
 


ing the subsidence will move both by the 2   H   H 
strike and dip, so the function of subsidence
probability F depends on all coordinates. 1   l  y  l  y 
Due to the independence of the influence of Y y    p p
 


individual coordinates, the equations: 2   H   H 
F  x, x , y , y , z , z  
(7) The function  from expression (9)
 F  x, x , t , t  F  y , y , t , t 
represents the standard Gaussian distribu-
represents the total subsidence probability tion curve.
for inclined seams [3].
2.4 Basic formulas for subsidence
2.3 Basic formulas for subsidence calculation of the inclined seam
calculation of the horizontal seam
If during the excavation of inclined lay-
When excavating layered deposits, the
used excavation methods that remove the ers, the exploitation area (  3 ) is of arbitrary
rectangular excavation panels of the layer, shape, the total subsidence M x, y, H  on
whereby the sides of rectangle are parallel the surface of the undermined massif based
to the coordinate axes- O by the seam on (7) will be,

No. 1-2, 2020 5 Mining & Metallurgy Engineering Bor

 If the seam dip angle (  ) and the seam
U x, y    F x, x , y, y, H , z  dx dy dz thickness ( d ) are constant, the area  3 is
3
prismatic with the base  2 (Figure 4).
(11)

Figure 4 Plane subsidence of inclined seam [3]

By introducing the new variables: x   If the sides of the excavation panel  2

; x   ; z   sin    cos  , the subsid-
ence calculation can be simplified, because
if   0, y    cos  and z    sin  ,
the difference when z  H is:
t  t  z  z   cos    y  y   sin  
 H  cos   y  sin 
and no longer depends on the variable  .
are 2a by the seam strike and 2l by the
seam dip, with the coordinate origin at the
intersection point of diagonals of the rec-
tangle it will be: a    a ,  l    l ,
the expression for calculating the subsid-
ence during inclined seams excavation is
obtained [3]:
U x, y   U 0 X x, y  Y  y  (12)

where:
1   ax   ax 
X  x, y    p p 
2   H  ctg  y   H  ctg  y 
   

1   bm y   bm y 
Y y   q    q 
2   H  ctg  y  
 H  ctg  y 


where: The subsidence curve will be the same

p0 in any profile by the dip, so it is a plane
p , q  1   sin  Q subsidence:
sin  ,
U y  U 0 Y y (13)
l cos  H  cos 
b , m where Y  y  is calculated from the formu-
1   sin  1   sin 
la (12).

No. 1-2, 2020 6 Mining & Metallurgy Engineering Bor

 In the case that   0 ,the formula (12)
is reduced to the basic formula (10) for cal-
culating the subsidence of horizontal seam.
2.5 Geometric interpretation of
parameters for the subsidence
predictive calculation
The subsidence calculation formulas
based on the stochastic method, formulate
two types of parameters:
- Parameters a, l , H , d ,  , b that are
known and determined;
- Parameters U 0 , p, m, q,  that are not
determined in the physical sense. The-
se parameters describe the behavior of
an undermined rock massif during
consolidation and need to be deter-
mined previously to find their relation-
ship with some unit that can be mea-
sured.
The following terms, definitions, and
symbols are adopted:
Dimensions of the excavation panel
a, l m are determined from the project
and depend on the exploitation area.
Seam depth ( H ) represents the mean
seam depth for each excavation panel. It is
calculated from:
H  HT  H S
where H T is the elevation of terrain over
the excavated seam [m] and H S is the depth
of excavation panel [m].
Figure 5 Parameters for mine subsidence prediction [4]
No. 1-2, 2020
Angle of seam dip  o varies. For the
predictive calculation of the movement
parameters, the mean values of the seam
dip for each excavated area are taken.
Maximum subsidence U 0 [mm] de-
fines the maximum possible subsidence
(Figure 5) that can occur at the full exca-
vation surface. Only one point M at the
ground surface has a maximum possible
subsidence. It is calculated from the ex-
pression [3]:
U 0  d  qu  i  cos  (14)
where: q is the rate of caving represents the
coefficient of subsidence of the hanging wall
and is taken as a real value based on the
analysis of former excavation. The most
commonly adopted value is q =0.7-0.9
which corresponds to the value that would
be taken based on an analogy from a basin
with the similar physico-mechanical proper-
ties of a hanging wall. The parameter q is
calculated from the expression:
q
f 2
300 sin 

exp 0,0027  0,0076 f  
where: f - is a strength coefficient of the
rock massif by Protodjakonov; i - coal
deposit recovery is determined with losses
g (%), which corresponds to the designed
excavation method, whereby i  1  g .
7 Mining & Metallurgy Engineering Bor
 Parameter p [m-1] depends on the re-
lationship between the side a and the hy-
pothetical element h . It is calculated from
the expression:
f 2
p . (15)
300  tg
Parameter m -to determine this parame-
ter, it is necessary to know the position of
the point with the maximum subsidence U 0
which is in the point M , with the ordinate
YM  m . The parameter m is in direct
connection with the angle of full subsidence
 (Figure 5), which is related with the an-
gle of draw  , ie. with auxiliary angles 
and  . Parameter m is calculated according
to the empirical formula:
m  H ctg (16)
Parameter b is not independent and is
calculated from the known data: the seam
dip  , dimension of the excavation field l ,
Figure 6 Angular parameters of undermined ground movement process:
 1 , 1 ,  1 - angles of the draw;  2 ,  2 ,  2 - safety angles;  3 ,  3 ,  3 - fracture angles;
 , 1, 2 - full movement angles;  - angle of full subsidence;  - angle of seam dip [5]
No. 1-2, 2020
seam depth H and parameter m (Figure 5).
It can be expressed through the angle  and
calculations according to an empirical for-
mula:
m
b  l cos   sin 
H
by entering the value m  H ctg it will be:
b  l cos   l sin  ctg , so the parameter
b is calculated from the expression:
sin   
bl (17)
sin 
Angles of draw ( 1 , 1 ,  1 ) are angular
characteristics of the moving process that
determine the ground surface area affected
by the displacements caused by excava-
tion, assuming that the deformations can
be neglected beyond that displacement
boundary. These are angles between the
horizontal and line joining the excavation
boundary in the pit with some boundary
point at the land surface (Figure 6).
8 Mining & Metallurgy Engineering Bor
 The boundary point is defined by the
value of measured displacement. Both
components of the total displacement vec-
tor (horizontal and vertical), which have the
millimeter values on the peripheral part,
can be measured. The boundary point is
easier to determine through the vertical
component –subsidence. The absolute val-
ue of the subsidence U g =10 mm is adopt-
ed for the boundary condition. For predic-
tive calculations, there is a problem of
adopting the draw angles. For that purpose,
Table 1 was created, in which the values
are given based on the interdependence of
angles: 1 , 1 ,  1 , [4].
The values of draw angles in the direc-
tions  1 and  1 are different. The sub-
sidence curve is asymmetric with U max
shifted in the direction the seam dip 
with the angle of maximum subsidence  .
When the coordinates of the points P and
Q (Figure 5):
y P  l  cos   H1  ctg
yQ   l  cos   H 2  ctg
are entered into the Eq. (13), the subsi-
dences in these points are obtained [3]:

U 0   b  m  l  cos   H1  ctg  
    q  b  m  l  cos   H1  ctg


UP    q 
2   H  ctg  l  cos   H1  ctg   H  ctg  l  cos   H1  ctg 
    

U 0   b  m  l  cos   H 2  ctg  
    q  b  m  l  cos   H 2  ctg


UQ    q 
2   H  ctg  l  cos   H 2  ctg   H  ctg  l  cos   H 2  ctg 
    

These are the general conditions for the massif inclined seams.

Table 1 Prediction values of draw angles (  1 , 1 ,  1 ) and maximum

subsidence angle (  ) [4]
Angle of draw Theory is valid, stable foot wall Theory is not valid, active foot wall
f a =0o a 0o a =10o a =20o a =30o a =40o a =50o a =55o a =60o a =70o a =80o a =90o
1 47 42 37 33 30 29 28 27 26 25
1.5  1  55 o 1 57 60 64 67 71 73 73 57 40 25
 85 81 78 76 74 74 73 75 80 90
1 53 47 41 37 33 32 29 28 27 26
2.5  1  60 o 1 63 66 69 71 74 75 76 59 41 26
 85 81 77 75 73 73 71 73 79 90
1 59 52 46 40 46 34 32 29 28 27
4.0  1  65 o 1 70 72 74 76 77 78 79 61 42 27
 84 80 76 73 72 71 72 70 77 90
1 62 54 47 42 37 35 33 32 31 29
5.0  1  70 o 1 73 75 78 79 81 83 82 64 43 28
 83 77 72 69 67 66 67 70 77 90
1 63 53 46 40 36 34 33 32 31 29
8.0  1  75 o 1 78 80 83 84 85 86 85 67 44 29
 82 76 71 67 66 65 66 69 76 90

No. 1-2, 2020 9 Mining & Metallurgy Engineering Bor

3 DEVELOPMENT OF THE GIS
BASED STOCHASTIC METHOD
FOR THE MINE SUBSIDENCE
PREDICTION
Since the subsidence prognosis using
equations of the stochastic Pataric-
Stojanovic method would be very compli-
cated and time-consuming in case of manual
data processing, the original computer pro-
gram package with the title MITSOUKO
has been created based on this method and
formulas. MITSOUKO is designed in the
Python programming language and provides
the mine subsidence and deformations cal-
culation, based on the simulating excavation
proces according to the adopted dynamics, at
any point of the land surface and represent-
ing the results owing to their integration and
further processing in the GIS. It consists of
three modules: PARAMETERS, SUBSID-
ENCE and DEFORMATIONS which repre-
sent the individual independent functions.
Each module starts with form according to a
textual description explaining its name and
function, the tags of input data that are loa-
ded (Read parameters) or computed (Calcu-
late parameters) into a particular module,
and the data values returned by a module
through the control loop. The modules have
control functions, descibing what the pro-
gram does according to the algorithm shown
in Figure 7.
Firstly, by entering the MITSOUKO
program, in the PARAMETERS module,
the function given by Eq. 9 is initialized [3].
Then, through the menu, the data obtained
based on geometric characteristics for each
excavation panel are entered: ID , dimen-
sions ( a, l ), seam depth ( H ), seam thick-
ness ( d ) and seam dip angle (  ). In the
next step, in a specified subroutine of the
PARAMETERS module, the following is
calculated for each panel: maximum subsi-
dence ( U 0 ), parameters ( m, b, p ), rate of
No. 1-2, 2020
caving ( q ), angles of draw (  ,  ,  ), and
angle of full subsidence (  ).
The local coordinate system is situated
symmetrically with respect to the first
excavation panel, with x - axis in the di-
rection of the seam strike, y - axis in the
direction of the seam dip and the coordi-
nate origin in the diagonal intersection of
this panel. The positions of all panels
( xi , y i , Fi ) are determined with respect
to the defined local coordinate system.
During the calculation for each panel, it is
necessary to rotate its coordinate axes for
the value of angle Fi (expressed in de-
grees) to the direction of axes of the local
coordinate system.
Subsidences are calculated in the SUB-
SIDENCE module. A certain subroutine
allows entering the coordinates of points in a
grid of a given density, through the assigned
distances between points ( x, y ), in the x
and y axes directions of the local coordi-
nate system. In this way, it is possible to
define the calculation limits for all panels up
to a limit subsidence value of 10 mm. Fur-
ther, the subsidence values after mining each
panel are calculated cumulatively, according
to the projected mining dynamics.
The individual results for each panel can
be exported (write) in tabular form in an
Excel file, which is also the preparation for
graphical presentation and spatial analysis in
GIS.
The DEFORMATIONS module at the
given points of the defined excavation pa-
nels calculates and presents in tabular form
in an Excel file the x, y components of the
tilt, and the directions of the main tilt due to
10 Mining & Metallurgy Engineering Bor
the mining out all panels. First, the compo- slope components of the individual panels.
nents of the main tilt, formed by mining one Based on these components, the main tilt
panel, are calculated in the direction of the and its direction are calculated. The program
absolute coordinate system, and then for all provides calculating curvatures for the given
panels, as in the subsidence calculation. The profile points in tabular form in an Excel file
components of the total tilt in the direction and their graphical presentation in the direc-
of the absolute coordinate system axes at a tion of any section after mining out each
given point are obtained by summing the panel.

Start

PARAMETERS
Input excavation field geometry
with subsidence parameters

Initialization
Read parameters: working panel ID; Read parameters: maximum possible
of function
dimensions of working panel ( a, l ); subsidence ( U0 ); parameters ( m, p ); rate of
caving ( q ); angles of draw ( ,  ,  ); angle of
t 1 2
s
 t  
2
seam depth ( H ); seam thickness ( d );
2
  e 2 ds
full subsidence ( ); point number along grid
coal seam depth (  ); positions of the 0

working panel in the local coordinate line; grid line number; point interval; line
system (xi,yi,Fi1) interval in x and y; point interval; x,y
(xi,yi,Fi1) x,y
x,y coordinates
coordinates coordinates start point of gride line
of
of dept
dept panel
panel vertex
vertex

Loop for working panel Loop point on the grid line and grid line

Array of x,y coordinate of panel

vertex transform Array of x,y points
Working panel number ID coordinates

Confirm x,y coordinate Generation of surface

for validation grid points

SUBSIDENCE
Subsidence components calculation
based on stochastic prediction method

Array result of subsidence components

DEFORMATIONS
Calculation of components:
tilt, curvature, horizontal displacement
and strain, based on stochastic
prediction method

Write the result at Excel file:

vertical displacement or subsidence;
tilt, curvature, horizontal displacement, strain

i=1,2,3, ... n Next

working
panel

End

Figure 7 Stochastic prediction algorithm for the subsidence and deformations

calculation using the MITSOUKO software

No. 1-2, 2020 11 Mining & Metallurgy Engineering Bor

 The program further calculates the
curvature radius, as the reciprocal value of
curvature in the interval, after mining each
panel. Also, the values of horizontal dis-
placements and strains caused by mining
all panels are calculated and tabulated.
GIS is used for creatinh complex geoda-
tabase, converting numerical data, imported
from the SUBSIDENCE and DEFOR-
MATIONS modules, in feature classes and
graphical data as well as for performing the
spatial analyses of subsidence and defor-
mations [7-11]. Since the pit consists of mul-
tiple panels with complex geometry, spatial
analyses requires a long time without GIS
because for each panel subsidence or defor-
mations must be presented cumulatively.
Implementing the stochastic method for
spatial analyses of subsidence and defor-
mations in the GIS is performed in three
steps:
The first step, Data module, involves
creating a geodatabase of the coal mine in
ArcCatalog, within the ArcGIS application
(ESRI Software Inc. http://www.esri.com/
software/arcgis/) with feature classes, tables,
and rasters. The feature class is a set of ho-
mogeneous spatial attributes in the form of
digitized vector data, in the same National
Coordinate System (MGI Balkans7). In or-
der to integrate the feature classes themati-
cally and spatially into the mine model,
within the given excavation panels, feature
datasets have been created, in which all
types of feature classes are entered. Feature
datasets with the feature classes related to
the spatial geometry: terrain topographies,
buildings, mining facilities, old mining
works and new exploitation area in pit, ex-
cavation panels with mining dynamics, ac-
tive and old mining premises, and geological
interpretation of the coal seam are created in
geodatabase of mine. Outside the feature
datasets, tables with subsidence, tilt, curva-
ture, horizontal displacement and strain,
No. 1-2, 2020
calculations from the SUBSIDENCE and
DEORMATIONS module in the MIT-
SOUKO program (Figure 8), rasters for the
subject area in the form of orthophoto,
geographic maps, situational plans of
mine and photographs are imported in the
geodatabase. Feature Datasets have been
created, in which all types of feature clas-
ses are entered. Also, using ArcMap, an
integrated part of the ArcGIS software
package, to create layers for displaying
feature classes from the ArcCatalog
(ESRI) is included. All tables of the sub-
sidence calculations have been transferred
from the geodatabase coal mine to the
ArcMap.
The second step, Subsidence module,
involves using the Display XY Data
command, the selected table of the exca-
vation panel, which contains the x, y
coordinates of the points and the calculat-
ed subsidence values, is added as a new
layer in the Table of Contents. Thereby, a
new feature class, panel ID_events, was
formed, which for panel contains many
points with values of x and y coordi-
nates and associated subsidences, georef-
erenced to the adopted coordinate system
MGI Balkans7. Following the same pro-
cedure, new feature classes were created
for all panels. The created feature classes
contain x and y coordinates of all points
in the grid 10x10 m with associated sub-
sidence values. The Spline interpolation
method from the Spatial Analyst Tools
palette is then used to create new layers
with contour subsidence lines for each
panel, by the cumulative subsidence trans-
formation from the previous to the new
state. This provides successive following
of the subsidence process on the map at all
mining stages. Digitization and formation
of feature classes for all panels in the
global MGI Balkans7 coordinate system
are performed in a map.
12 Mining & Metallurgy Engineering Bor
 Data module
GIS layers (terrain topographies, buildings, geological interpretation of the
coal seam, mining excavation map, working panels, etc.)

Topomap
Feature datasets
(feature classes) Streets map
Plans
Geomaps

Surface point layer Ortophoto

GEODATA RASTERS
BASE Strain
Polygon panel layer
Horiz. displec
Curvature

Tilt
Subsidence
MITSOUKO
TABLES

Subsidence module

Geometry GIS spatial Subsidence grid-points

working panels analysis in new feature class

Subsidence spline
interpolation

Contour
subsidence lines for
working panels

Deformations module
Calculation
GIS spatial Deformations
Curvature
analysis Tilt/Curvature

The radius of curvature Contour tilt lines

Curvature diagram Diagram for main tilt

Result output

Figure 8 Algorithm connecting stohastic prediction analysis with

the GIS spatial analysis function

No. 1-2, 2020 13 Mining & Metallurgy Engineering Bor

 Figure 9 Interpolation of subsidence contour lines after mining excavation panels

In the third step, Deformations mod- The Spline or Kriging interpolation

ule, based on the subsidence values for
individual panels obtained in the GIS spa-
tial analysis, tilt is calculated cumulatively
as subsidence increment after mining each
panel.
method from the Spatial Analyst Tools
palette is then used to create the new la-
yers with the contour tilt lines (Figure 10).
It is possible to draw the tilt diagrams and
the main tilt according to a given profile.

Figure 10 Interpolation of tilt lines after mining out all excavation panels

No. 1-2, 2020 14 Mining & Metallurgy Engineering Bor

 Based on the subsidence increment, after
mining each panel, for the profile points, the
curvature is calculated and presented, i.e. the
curvature radius as its reciprocal value.
4 CONCLUSION
The calculation of subsidence by the sto-
chastic Pataric-Stojanovic method and origi-
nally developed MITSOUKO program
package integrated with the GIS are power-
ful tools for predicting subsidence and moni-
toring the impact of underground mining
works on the land surface.
Benefits of this approach are:
- reliability of the method for calculation
of subsidence and deformations for in-
clined seams;
- equations for inclined seams, in cases
when   0o , give solutions for hor-
izontal seams;
- definition of draw angles in terms of
the possibility of their application
when seam thickness and excavation
depth increase, and
- mutual connection between the angular
parameters of the displacement pro-
cess.
REFERENCES
[1] Kratzsch H., Mining Subsidence
Engineering, Springer-Verlag, Berlin,
Germany, 1983.
[2] Litwiniszyn J., The Theories and
Model Research of Movements of
Ground Mass. Proceedings of the
European Congress on Ground
Movement, Leeds, 1957, 202-209
[3] Patarić M, Stojanović A., Movement
of Undermined Terrain and Protection
of Structures on Mining Activities,
University in Belgrade, MGF,
Belgrade, 1994 (in Serbian)
No. 1-2, 2020
[4] Đorđević D, Vušović N., Forecast
Calculation of Movement and
Deformations of the Undermined
Terrain, University in Belgrade, MGF,
Belgrade, 2014 (in Serbian)
[5] Đorđević D, Vušović N, Ganić A,
Svrkota I., Angular Parameters of
Undermined Ground Movement
Process in Unknown Areas, University
of Belgrade, Faculty of Mining and
Geology, Underground Mining
Engineering, 19 (2011) 125-136.
[6] Burrough PA, McDonnel RA.,
Principles of Geographical Information
Systems. 2th ed. Oxford, London, 2006.
[7] Jiayuan C., Fengshan M., Jie G., Rong
L., Guowei L., Assessment of Mining-
Related Seabed Subsidence Using GIS
Spatial Regression Methods: A Case
Study of the Sanshandao Gold Mine
(Laizhou, Shandong Province, China),
Environmental Earth
Sciences 78(1)(2019) 26-37
[8] Malinowska A, Hejmanowski R.,
Evaluation of Reliability of Subsidence
Prediction Based on Spatial Statistical
Analysis, International Journal of Rock
Mechanics and Mining Sciences 46
(2009) 432–438.
[9] Blachowski J., Application of GIS
Spatial Regression Methods in
Assessment of Land Subsidence in
Complicated Mining Conditions: Case
Study of the Walbrzych Coal Mine
(SW Poland), Natural Hazards, 84
(2016) 997-1014
[10] Petrović V, Borisov M, Zlatanović D,
Djurdjevac Ignjatović L., Vertical
Accuracy of Data on the Topographic
Maps and their Application in Mining,
Mining and Metallurgy Institute Bor
No 3-4 (2017) 39-50
15 Mining & Metallurgy Engineering Bor
[11] Petrović V, Zlatanović D, Borisov M,
Djurdjevac Ignjatović L., Concepts of
3D Terrain Modeling and Geomorpho-
metric Analysis in Mininig, Mining
and Metallurgy Institute Bor No. 3
(2016) 1-12
No. 1-2, 2020
[12] Vušović N, Vuković M, Svrkota I,
Stojković P., Possibility for Appli-
cation of GIS Technologies in RTB
Bor Group, Mining and Metallurgy
Institute Bor No. 3 (2016) 21-30
16 Mining & Metallurgy Engineering Bor